tl;dr– The question was answerable because someone could post the information necessary to help the asker figure out what they were asking about. While this may've meant telling the asker that they had some investigation to do, that's perfectly fine – just like it's fine to give abstract answers on SE.Math or instructions on how to figure out a computer problem on SE.SuperUser.
Somewhat belated, but wanted to explain my reason for voting to reopen...
Discussion: Good answers are maximally concrete, but sometimes that's still pretty abstract.
Hypothetically, say someone on SE.Math asks:
If x + y = 2 and x = 1, what's y?
Then we'd tell them that y = 1.
But say someone asks:
If x + y = 2, what's y?
Should we close their question for not having enough information? Or, is it okay to be abstract in saying that y = 2 - x?
I'd frame this as an issue of folding abstract syntax trees. When we answer questions, we:
Parse the question.
Fold it as much as possible.
Post the result as an answer.
For example, when answering either SE.Math question, you'd probably arrive at the fact that y = 2 - x. However:
In the first case, you continue to find that y = 1 because you're able to.
In the second case, you stop and post y = 2 - x because that's as far as you can go.
Point being that questions are still answerable even if we don't have enough information to "fully" answer them.
The SE.Academia question discussed in this SE.Academia.Meta question could be answered abstractly.
With respect to this SE.Academia question, it'd seem hard to give a concrete answer (like y = 1) to why the asker's female students gave them lower scores than their male students.
Still, it seems like we could answer it.
I'd suggest something like this:
Put yourself in the asker's shoes.
Imagine how you'd find the answer to this problem.
Fold it as far as possible.
Post the answer.
Of course, there'd be a lot of ways that you could, in theory, find an answer to the question.
For example, one possible answer might be:
Invent a time machine.
Go forward into the future to where there's thought-reading technology.
Bring that technology back to the present.
Use it to figure out what the students' reasonings were.
That'd be sorta like telling the SE.Math person that they should get a quantum-computer to find an approximate solution for y. Which, obviously, would be a bad answer despite technical correctness.
Instead, good abstract answers ought to be reasonably implementable. For example, a good answer might:
Present the space of likely explanations based on published researched or/and personal experience.
Suggest a practical methodology for narrowing down the presented possibilities to as few as possible (ideally one).
For example, a good answer might be like:
Studies have shown that, when there's a gender disparity in student feedback, it's likely due to one of the following reasons:
Teaching style appealed more to one gender than the other.
Subject was more interesting to one gender than the other.
Students perceived instructor as having been sexist.
In order to determine which of these common explanations may be applicable to your case, you should:
Perform this inventory to determine if your teaching style has a gender bias.
Check this table for gender preference statistics on your field, and then this correlation to estimate the expected effect on student feedback.
Consult with your TA's to get their opinions on if perceived sexism may've been an issue.
Of course, if the asker had included more information, e.g.
a detailed description of their teaching style;
a full description of their subject and course content;
feedback from their TA's on if there may've been perceived sexism;
then instead of posting the more abstract answer, we could fold it into a more concrete conclusion.
The point's just that we don't need to simply say something like
Your female students were more bored by the subject matter than your male students.
if we don't actually know that to be the case.
Conclusion: The question was answerable, even if not concretely.
In short, while there may not have been enough information to precisely explain the gender disparity that the asker saw in their student evaluations, a good answer wouldn't need to provide such a concrete answer any more than a good SE.Math answer would need to provide a specific number.
Instead, it's okay to give an asker a framework that they can use to find their concrete answer. For example:
If someone's asking for a solution to a math problem, it's okay to give an algebraic response instead of a number.
If someone's asking how to fix their computer, it's okay to give them instructions on how to diagnose the problem before actually telling them how to fix it.
If someone's asking about how to interpret student feedback, it's okay to give them instructions on how to go about examining that feedback.